The invention relates to apparatus for measuring the A.C. characteristics, including, for example, the resistance, conductance, and impedance of a circuit element to which both an A.C. test signal and a D.C. test signal are applied simultaneously.
FIG. 2 depicts an example of the component measurement apparatus known in the prior art for performing automatic measurement of the A.C. characteristics of a component with both A.C. and D.C. signals applied simultaneously. The purpose for providing D.C. signals concurrently with A.C. signals has been to obtain the characteristics of the component under test in a simulated operating environment by measuring its characteristics at several levels of D.C. bias and over a range of A.C. frequencies. To make such a measurement, a D.C. bias voltage with an accurate, adjustable value must be superimposed on the A.C. signal applied to the component under test.
The circuit in FIG. 2 shows an example of measuring the characteristics of diode 3, which can be either a schottky diode or a PN-junction diode. In FIG. 2, A.C. power source 1 is connected through a source resistor 2 to one side of the sample diode 3, at terminal A1. The other side of the sample diode 3 is connected at terminal A2 to measurement terminal A6 and to the inverting input of an operational amplifier 5. Measurement terminals A3 and A4 are respectively connected to terminals A1 and A2 at either side of the component under test, sample diode 3. A variable range resistor 4 is connected between the output of operational amplifier 5 and its noninverting input. Measurement terminal A5 is connected to the output of operational amplifier 5. The noninverting input and ground terminal of operational amplifier 5 are tied together and are connected through a capacitor 8 to A.C. power source 1. Monitoring terminal 10 is connected to the noninverting input of operational amplifier 5 through a capacitor 7 and to terminal A1 through resistor 6.
The D.C. bias voltage is supplied by a variable D.C. voltage source 11 connected through transformer 9 across capacitor 8. The positive side of voltage source 11 is connected through the primary winding T1 of transformer 9 to the side of capacitor 8 connected to A.C. source 1. The negative side of variable voltage source 11 is grounded, and the secondary winding T2 of transformer 9 is grounded and connected to the opposite side of capacitor 8.
To make measurements automatically over a range of A.C. frequencies, the amplitude and frequency of A.C. power source 1 and the resistance of variable resistor 4 are automatically controlled by a control circuit (not shown).
In the circuit of FIG. 2, the A.C. current from A.C. power source 1 flows through source resistor 2, diode 3, variable range resistor 4, operational amplifier 5, capacitor 8 and then back to A.C. power source 1.
The D.C. current from voltage source 11 flows through primary winding T1 of transformer 9, A.C. power source 1, source resistor 2, diode 3, variable range resistor 4, operational amplifier 5 and secondary winding T2 of transformer 9 to ground. The D.C. volatge of the terminal A1 side of diode 3 is measured by connecting a voltmeter to monitoring terminal 10. Resistor 6 and capacitor 7 shunt any A.C. signal and enable terminal 10 to measure only the D.C. voltage.
The frequency-conductance characteristics of sample diode 3 at the selected D.C. bias voltage are measured as follows. First, the output voltage of D.C. power source 11 is allowed to stabilize at the selected value to provide a constant D.C. voltage to diode 3, measured at terminal 10. Then an A.C. signal at the desired frequency is provided from A.C. power source 1. During such initialization, the appropriate value of variable range resistor 4 is automatically selected. The conductance of diode 3 at the applied frequency can be obtained by measuring the A.C. voltage drop across measurement terminals A3 and A4 and across measurement terminals A5, A6. Similarly, the frequency-conductance characteristics of sample diode 3 at various frequencies can be obtained by sweeping the output frequency of A.C. power source 1.
However, the D.C. voltage drop across the diode 3 may not correspond to the D.C. bias voltage measured at terminal 10 or to the voltage output of source 11. The measurement at terminal 10 does not take into account the voltage drops across resistor 2, FIG. 3 is a plot of the D.C. voltage and D.C. current across diode 3 versus the output voltage of D.C. power source 11 when the measurement apparatus of FIG. 2 is used to apply a D.C. bias voltage to diode 3. As FIG. 3 shows, the D.C. current through diode 3 increases as the output voltage of D.C. power source 11 increases. The voltage drop across source resistor 2 increases as the current in the circuit increases and this leads to unnegligible errors in the conductance values measured versus applied D.C. voltage if the D.C. voltage is not measured at monitoring terminal 10. When the output voltage of D.C. power source 11 is 1.0 V, this error is 20% because of the measurement error caused by the voltage drop across source resistor 2.
An additional source of error is the voltage drop across the input terminals of operational amplifier 5. Since the amplification factor of amplifier 5 cannot be infinite, the input impedance (approximately equal to the resistance of vaiable range resistor 4 divided by said amplification factor) cannot be zero, so a further unnegligible voltage drop occurs. The D.C. input impedance of operational amplifier 5 and the D.C. voltage drop across the input terminals of operational amplifier 5 also fluctuates with changes of the resistance of vaiable range resistor 4. These effects can lead to fluctuations in output voltage at terminal A2.
FIG. 4 shows the effects of the afore-described errors on the accuracy of a conductance measurement made on diode 3 (with the output voltage and output current of D.C. power source 11 maintained at 0.5 V and 1.0 A). The solid line curve in FIG. 4 shows the measurements made with the apparatus of FIG. 2. Apparent from this plot is the significant variation in the measured conductance value caused by changing the resistance of variable range resistor 4. In measuring the doping characteristics of a diode, a very accurate measurement of the frequency-conductance characteristics and bias voltage-capacitance characteristics of the diode is necessary. The measurement apparatus of FIG. 2 could not be used for such measurements because of its inherent relatively large measurement error. The connection of a voltmeter can be arranged to monitor D.C. voltage across diode 3. However, whenever the resistance of range resistor 4 is changed, the output voltage of D.C. power source 11 must be adjusted such that the voltage across diode 3 is the selected value.
It is a principal object of the present invention to provide measurement apparatus for obtaining various A.C. characteristics of electronic components at selected D.C. bias voltages which can be accurately maintained at the selected values across the component under test.
In accordance with the illustrated preferred embodiment of the invention, a control circuit accurately maintains the D.C. bias across the component under test. The component measurement apparatus is comprised of a D.C. power source to apply D.C. bias voltage to the component under test, an A.C. power source to apply A.C. signals to the component undr test concurrently with said D.C. bias voltage, and a control circuit to detect the D.C. voltage across said sample and to maintain the D.C. voltage across the sample at the selected value while the measurement is made.
In accordance with the present invention, a stable D.C. bias voltage is maintained across the component under test by sensing the difference between the output voltage of the D.C. bias voltage source and the voltage drop across the circuit element under test, to control the D.C. bias voltage applied and to compensate for unwanted or variable voltage drops in the measurement circuit. The apparatus of the invention compensates for both fixed resistance in the circuit and for variable resistance due to changing the range resistor. Furthermore, an ammeter can be included in the D.C. bias circuit without adversely affecting the accuracy of the measurement.